Structural Damage Identification For Bridges Based On Wavelet Analysis

It is well known that the bridge is a key component in the transportation system. With the flourishing development of China's transportation, the number of bridges increases rapidly and the safety evaluation becomes much more important. The bridges can be damaged gradually by many adverse factors such as environmental erosion, material aging, excessive use and so on. Under the influence of those factors, the damage will be more and more serious. Finally, it will cause bridge collapse and even other major incidents which are great threats to the safety of people's lives and property. In order to prevent frequent occurrence of these incidents, we must identify and locate the potential damage as early as possible. Meanwhile detecting and evaluating long-term healthy on the structure of important bridges are also crucial. Fortunately, the detection system of many large bridges had collected a lot of deflection data during the service period, which can characterize the deviating level of bridge observation point from the equilibrium position, then we can find the internal damage by processing and analyzing these deflection data.Wavelet analysis is developed in recent decades. It's considered as a breakthrough of Fourier analysis with the excellent ability of time-frequency analysis. Processing and analyzing of measurement signals are the key issues of identifying the bridge damage in structure. By using wavelet transform, the signals of bridge structural health monitoring system are described on different scales. This approach is based on the assumption that rubbing produces features in a monitoring signal which result in very large wavelet coefficients at a certain level can detect the presence and occurrence of damage.At present, research of Wavelet analysis in structural damage diagnosis basically remains in the numerical simulation stage, and there are few practical applications. Therefore, applying wavelet analysis of bridge health monitoring to the practice needs more further research. So this thesis puts forward an identity strategy based on Wavelet analysis in which we can identify bridge damage in structure by analyzing the data of measurement signals from multi-scale.The main work of this article:—Reviewing the concept and significance of the bridge health monitoring, summarizing the research status both at home and abroad, and introducing some relevant knowledge of Wavelet analysis. —Making effective noise reduction for large quantity of raw data and getting structure parameters directly using wavelet analysis,identifying the structural damage and warning timely in the principle of wavelet singularity detection.—Processing and analyzing the measured deflection data of Shi Banpo Bridge using the damage diagnosis theory based on "Energy Damage". The energetic feature vectors of wavelet packet node are proved to be more suitable to evaluate the damage detection of bridge structures, for it is sensitive to minor damage.Generally speaking, this thesis is focused on the identification and location of the bridge structural damage with the help of signal processing methods of Wavelet analysis, damage warning and structural damage feature parameter extraction. .